Retinoids prevent and reverse the process of neoplastic transformation and inhibit tumor promotion. However, systemic toxicity and teratogenicity have hindered the development of natural and synthetic retinoids as useful drugs for the chemoprevention of cancer. With the identification of the family of nuclear retinoic acid receptors (RARs) and the discoveries of their effects on gene expression, the mechanism of retinoid action at the molecular level can now be investigated. We propose to develop receptor- specific retinoids and antagonists for chemoprevention of cancers induced in epithelial tissues. Correlations of the results of retinoid transcriptional activation of and binding affinity for specific RARs with retinoid structure and the ability to regulate cell differentiation in more complex systems are invaluable for the efficient design of retinoids having enhanced chemopreventive effectiveness. In this Program Project comprising four integrated Research Projects and a Bioassay Core, we propose this approach, namely: (1) design and synthesis of retinoids based on structure- activity correlations (Project I), (2) theoretical studies involving structure-activity and multivariate analyses of bioassay results to define the optimum pharmacophore for each receptor (Project II), (3) investigations of how nuclear receptors mediate the retinoid response and evaluation of retinoid functional activation of the RARs (Project III), (4) X-ray crystallographic investigations of the three-dimensional structures of the receptors to define the geometry of the ligand-binding domain for the RARs (Project IV), and (5) determination of RAR binding affinity, and evaluation of the activity and cytotoxicity of receptor-selective retinoids in epithelial cell culture assays (inhibition of anchorage-independent growth of JB6 epidermal cells and differentiation of SCC13 human squamous carcinoma cells), followed by assays for the inhibition of induction of the tumor marker ornithine decarboxylase in mouse epidermis by tumor promoters and toxicity in the mouse (Bioassay Core). The suitability of candidate compounds for further development as chemopreventive agents will be established using a decision network method. A team of recognized scientists and clinicians will serve as a Scientific Advisory Panel to help guide research Progress.